Risk for HIV infection in the United States falls along sharp lines of disparity related to sexual orientation and race/ethnicity. HIV infection has always taken a heavy toll on men who have sex with men (MSM) and also on African Americans. However, HIV incidence disparity is most striking of all for MSM who are also African American, a segment that probably accounts for less than one percent of the American population but fully 25% of all new HIV infections in the United States. Surprisingly, very few HIV prevention interventions designed for African American MSM have been undertaken, and most existing approaches have relied on individual or small-group risk reduction counseling. These existing models have limited cultural and practical suitability for reaching high-risk African American MSM who are hidden in the community and unlikely to seek out sexual risk reduction counseling. Interventions directed to individuals alone also do not attend sufficiently to the important influence played by social norms and social network factors on risk behavior. In contrast to individual counseling models, HIV prevention interventions that are directed to social networks of Black MSM in the community and that work through natural influence channels within these networks hold the potential for reaching deeply into hidden and vulnerable populations of African American MSM. Network-level HIV prevention interventions rely on oral communication traditions and cultural models appropriate in minority communities, and can strengthen social network normative support for risk reduction. Following a 1-year formative research phase that will provide data needed to contextualize the study's recruitment, assessment, and intervention content and delivery methods, we will undertake a social network-level intervention outcome trial in 3 cities: Milwaukee, Cleveland, and Miami. Project-wide, 24 sociocentric networks of African American MSM-each expected to be composed of approximately 40 people (total n=960 participants)-will be recruited. 12 of the networks (n=480 participants) will be randomized to a comparison condition and receive HIV/STD testing, treatment, and individual counseling. The other 12 networks (n=480) will receive the same counseling and testing and also a social network intervention that identifies, trains, and engages empirically-identified influence leaders of each network to deliver ongoing, theory-based, and culturally tailored HIV prevention advice, recommendations, and support to other network members. The study will determine the comparative effectiveness of the social network intervention to the counseling, testing, and treatment comparison condition on measures of sexual risk, substance use associated with sexual risk behavior, and STD/HIV incidence assessed with laboratory measures at 6- and 18-month followup points. The study aim is to develop and determine the efficacy of an intervention modality capable of reaching and preventing HIV among high-risk African American MSM in the community.